[universeplastic]

www.universemolding.com

[universeplastic]

CNC prototype manufacturing services are very affordable
https://www.universemolding.com

[peteeng]

Hi G - You still have not answered the modulus question. This impacts the FE results considerably. I'll put this another way. There are several concrete and EG builds on this forum and other forums. Many of them have trouble once cast getting the rails right. For an accurate machine this is imperative. The builders have gone to great lengths with accurate moulds and casting care, yet the build has difficulties. Its very generous of them to publish the difficulties rather than hiding... Your machine is big and the scale introduces its own hurdles as well with casting. But proceed as planned and eventually the proof is in the eating as they say... brings me back to Italian ice cream memories... Peter

[G-Spot]

Hi G - You still have not answered the modulus question. This impacts the FE results considerably. I'll put this another way. There are several concrete and EG builds on this forum and other forums. Many of them have trouble once cast getting the rails right. For an accurate machine this is imperative. The builders have gone to great lengths with accurate moulds and casting care, yet the build has difficulties. Its very generous of them to publish the difficulties rather than hiding... Your machine is big and the scale introduces its own hurdles as well with casting. But proceed as planned and eventually the proof is in the eating as they say... brings me back to Italian ice cream memories... Peter

1.936e+10 Pa

My philosophy on cement is if you are having problems using it, then it's usually because you are not using enough. My floor slab blah blah...yeah when you are trying to cast large thin flat slabs of it. Once you get more than 200mm thick it is forever. Built my own home with it, I originally used precast pieces until I discovered how crap the concrete they use really was compared to what you can make yourself, its always about the money, trying to build faster for the least amount of money and skill, like I said earlier cement with the correct amount of water in it looks nothing like what you see used commercially, it's difficult to work with and without a proper vibrator almost impossible to cast with. If you told me an experienced concrete layer was having problems making concrete machine bases I might be worried. But it really is an art, where to drop the vibrator how long to keep it there, how you pull it out, every time I ever gave this job to a newbie they messed it up, always too much, too slow. The mistakes I see with these machine base builds are too much buggering around with the cement, get it in the mold, vibrate it quickly, then leave it alone. Never saw one using a proper concrete vibrator always some home brew vibrator thing, vibrating the formwork and not the cement, the usual suspects.

Cement and steel are like Romeo and Juliet anyone who says they don't belong together is a liar.

[jackjr-123]

Concrete for general construction has nothing to do with concrete for machine bases.
I think you are vastly underestimating the many potential issues when filling hollow steel structures with concrete.
The main issues are:
1. shrinkage (during curing and over time)
2. mismatch of thermal properties (mainly inertia)

Also proper UHPC made for machines doesn't need to be vibrated.

[G-Spot]

yet the build has difficulties. Its very generous of them to publish the difficulties rather than hiding. Peter

What difficulties are they having ?

[G-Spot]

Concrete for general construction has nothing to do with concrete for machine bases.
I think you are vastly underestimating the many potential issues when filling hollow steel structures with concrete.
The main issues are:
1. shrinkage (during curing and over time)
2. mismatch of thermal properties (mainly inertia)

Also proper UHPC made for machines doesn't need to be vibrated.

I also suffer from thermal inertia that is is why I chose to live somewhere where the temperature never varies more than 10 degrees day and night all year round. Is this an issue if you keep the machine in a room which is always the same temp?

[jackjr-123]

Watch this:
https://www.youtube.com/watch?v=_XGF2FZUTms

[peteeng]

Hi G - Sorry to harp on this and I'll make this the final summary comment on this subject from me - The grouts used for making machines do not need vibrating. In fact they strongly state that it should not be vibrated. They self compact and self de air. Grouts (to distinguish between portland concrete and other concretes) are compensated for drying shrinkage and plastic shrinkage. In USA they are called Class C concrete or double compensated.

The main difficulty is that the cast rail foundations are not flat or square and if they have used steel inserts to create the rail foundation they are not flat or square, post cast. The commercial makers that use grout or mineral castings post machine the foundations just like you would a metal machine, They have large wet grinders to do this. DIY people once cast can't move the machine to a mill big enough to correct the foundation. Many have to manually lap or scrape the foundation into spec and that's what I expect you may have to do.

Casting concrete has a couple of hurdles:
1) The process is very exothermic. The cast may get to 80C or higher. This temp delta means the mass grows due to thermal expansion. This moves things in your mould and then it sets bigger than intended. In your case the steel will expand. Then it cools and this shrinkage creates micro cracking if the concrete is not plastic enough in this phase and moves the inserts. In your case using the steel as an exoskeleton may work out, but it may go against you as well by restraining the cast. Restrained castings crack and/or disbond from the mould/exoskeleton
2) drying - portland cement has an excess of water in it to create a flowable product. This excess water desorbs from the mass over time creating more shrinkage and internal cracking. Cracked concrete has no tensile stiffness. Thats why they add steel fibres (or other fibres) to the mix to bridge the cracks
3) Your modulus of 19GPa is low so that's good (conservative). But the plywood I use is 20GPa (by test) so you may as well use mass plywood if it contributes so much to your stiffness. I've made many composite materials (mineral castings) at 20GPa and have decided not to use them as they are too low in modulus.
4) I'm concerned that coupling of the steel and concrete over time will not exist. Concrete machines use lots of embedded mechanical connections between the steel and the concrete. eg steel reinforcement is not smooth but it is chevroned with scales. This is to create a mechanical key vs a bond
5) steel is 200GPa and concrete is 30GPa so strain will prefer to travel in the steel vs the concrete. Granted the FE model works but for many reasons in the real world it may not over time
6) your about to put alot of $$$ effort and time into this and all of that needs to be well spent. Often material cost is not the main $$$ of the project but the material is the heart of the machine. You get what you pay for...

Nothing ventured nothing gained and sorry to be devils advocate - some stories do have a happy ending... which is what I want for you and others... Peter

On the issue of exothermic expansion - I vacuum cast fibreglass and carbon fiber slabs, some times steel fibre slabs. When they are say 20mm thick or more they get up to 80C when curing. They are around 55-60% by volume fibres so theres not much epoxy in them. If I use an aluminium plate for the flat mould the Al grows and distorts in the middle. When cured and released the composite plate is cupped. Some times by 2mm over say a 500mm slab. So steel would do the same but half the expansion. I then have to machine that plate flat if it has to be flat... If I use formply as the mould it's even worse. So I expect that the strain created by exotherm in a thick casting is quite a bit. The concrete chemist I consulted with for a time said that for thick castings say a commercial floor they use a "slip" layer of plastic in the bottom of the form so the concrete can expand then shrink without cracking...

To solve my flatness issue the mould has to be prewarmed to the exotherm temp but then that creates issues with too much energy available for the epoxy and it can over heat and burn. Same issue with concrete if it overheats it damages itself as well... I use plastic in the bottom now as a slip, still cups a bit as the wax under pressure still has a very good grip on the cast....

So I have spent some time identifying suitable grout for this application. It has its place but same as all the other materials has hurdles to jump and characteristics to understand. My main issue is that the quoted modulus of commercial concrete, UHPC and grouts has not been able to be replicated by myself. They test these materials in compression and I think the value is biased in the test. I test in flexure as machine parts are "bent" in practice, like your Z axis. I expect your E=20GPa is realistic, but a microcracked mass has zero stiffness (except in compression) .... sorry I'm a pessimist when it comes to some subjects these days ...

[G-Spot]

Went back to the drawing board, clean sheet.

I wanted to work out geometrically and structurally what was the strongest shape I could fit in the space available. Went and measured my workshop roof, which has an up and over door that restricts the ceiling height to 2.4m.

Created a structure that keeps as much of its material away from the centre of inertia.

I also realised from the last design that for plus 50N/um stiffness at that scale you will need at least 2000 Kg of steel.

I cant afford 2000 Kg of new steel, so I needed to build this machine from scrap steel. The only scrap steel that is plentiful that could do the job would be I-Beams, thick plate steel is not available at scrap yards, when people buy the stuff new they make things from it, they don't store it.

So I downloaded a list of easily available I-Beams to see what I could do.

Ladies and Gentleman the OctaBeast returns!!!!!

The same weight as my previous design but with double the stiffness and capabilities. More than 100N/um in all axis.

It's modular and all the eight pieces and plates etc are all the same angles cut lengths etc, so any jigging and machining are quick to do.

[G-Spot]

You can upgrade it endlessly.

You can put a large bed on it and use it to machine sheet or plate. Or you can put a small bed on it and use it to machine upto 500mm Z axis.

I am already gearing up to make the frame.

Purchasing 3 lasers and beam splitters to make a device capable of aligning all three axis together.

Will use optical grade glass and all 3 axis will still cost less than a SINGLE bearing block.

Already done most of the research necessary for the laser beams, going to take them to the dam or beach and align them with the water, yes I know I am going to see the curvature of the earth, but I am only going to use the water or beach as place I can shine the beams uninterrupted for long distances. Like 5km. Not to use the surface of the water for anything.

[G-Spot]

The main difficulty is that the cast rail foundations are not flat or square and if they have used steel inserts to create the rail foundation they are not flat or square, post cast. ...

The answer is pre-stressing and symmetry. This is how it is done with concrete.

You have a bar of metal close to the surface at the top, and a bar of metal close to the bottom.

You pull them apart with equal force, it will need to be 2 trees if it's steel, it needs to be up there near the yield strength of the material you are using.

You then poor the concrete into your mold and wait however many days it take for the concrete to reach more than 90% of it's strength. You then cut them away from the trees and you wont have any distortion.

You keep saying that the steel in reinforced concrete is doing the strength part while the concrete is sitting there along for the ride, this is incorrect the concrete does all the tensile strength part until it cracks and only then does the steel take over. In pre-stressed concrete, not post stressed concrete, the steel is already doing the work and the concrete is helping out because it is being held in compression by the steel and concrete works well when it is held that way.

You will know if you are doing it right because you will be looking for 2 trees or enormous pieces of buried concrete to do it, You will pull your house down if you try to use the walls.

I also own a lab grade vacuum machine and have done vacuum infusions. Also used to own a hydraulic intensifier that used to be on the metal press of the local Toyota factory, that would boost the 300 bar from a standard power pack up to 1000 Bar. So no stranger to the extremes of pressure.

[peteeng]

Morning Spot - There are two main philosophies in structures: monocoques and space frames (or 3D trusses). Monocoques, such as you are designing for minimum weight are limited by local buckling and large surface areas which contribute to their limit weight. Spaceframes depend on the triangular elements and the global size of the network for their global stiffness. The element itself is not dependent on its geometric inertia as the loads are always axial. So its area is important, so it can be a solid rod and optimised easily on strength or compression stability. If the element is compression dominant then it can be hollow to help with limit buckling maybe. Your Octobeast design looks to me to be a candidate for a spaceframe vs a membrane structure. I have played with spaceframe gantries and they are lighter but are more complex to make. You don't seem to be worried about complexity so maybe have a think about a spaceframe. I'm sure scrap barstock is out there.... A tower crane is an example. If it was made as a monocoque or membrane structure it would be very heavy due to its surface area but as a space frame it works. Peter

[G-Spot]

I feel confident to build this without too much hassle. The reason being is you only have to make one perfect jig and then it will work.
I always make jigs, they always work really well. Any machine consisting of hundreds of different parts is the one you will battle to maintain quality.
With a jig and identical parts each one you make gets better.
Get the end plates cut and drilled accurately so they are all identical. Order an extra 2 for the jig.
After you make the jig, you bolt the end plates to each side of your jig then tack the I beams in.
Then you assemble it and make all the final welds with it bolted together, it is in itself the strongest jig you could ever make.

I purchased the best featured TIG welder that can run on a single phase for this project, 250A AC/DC/Pulse going to rig it up semi CNC assist, with an auto wire feeder and tractor wheel,
that way I can make all the welds very quickly with each one having the same amount of heat pumped in.

That's the first CNC thing I am going to start making immediately. Going to use the all metal wire feeder with stepper motor that the 3D printer guys fit as an aftermarket mod, the ones most welders use are plastic and normal motor with no positional control. With TIG you don't want to just keep feeding it in, you want to be able to dab it like when you are doing it manually. With a stepper or servo wire feeder you can make the wire move a few mm back from the weld pool.
Then I might use the same motor and mechanics for the tractor wheel which is just going to be a metal or silicone wheel fitted to the torch that spins when you are welding, I can rest the weight of the torch on the wheel and allow it to pull the torch forward.

Any recommendations of what welding wire to use in the wire feeder? it's diameter limited to what the 3D printers use but last time I checked it wasn't a problem, I think there was a welding wire that will go through it. Don't want to even try TIG welding feeding the wire continuously into the weld.

Then the next ability I have to be able to demonstrate before purchasing any beams or steel is being able to accurately grind the surfaces flat and at the required angles.

Thinking of purchasing a large saw like in the pics below to cut the beams to within a few millimeters of the correct dimensions and then I am going to make my own surface grinder just to grind the surfaces for this job. Done a few days research already with surface grinders, going to go for whatever I can cobble together controlled by an Arduino and lasers. Already have the code, it's the same as what they use for Astronomy to detect planets slightly dimming their stars as they pass in front of them.
The laser is scanning the surface you are grinding and by the intensity of light can detect how flat it is.

All the laser stuff I have already made in the past, works really well, all the algorithms and Arduino code is already out there because of the Astronomy guys needing it to make their mirrors and flat surfaces.

So yeah I just have to cobble together the surface grinder from hell and use the Arduino to keep it accurate.

Going to get 3 lasers and fix them with adjustable mounts to a piece of thick glass or surface plate, then align them so they each produce a beam for the X Y Z axis, will do this at the closest dam which has a minimum distance of 2km.

The best thing is all the laser and optics stuff is as cheap as buttons and I have some really nice lens and lights and lasers.

[joeavaerage]

Hi,
I worked for seven years fixing and servicing welding equipment.

The biggest issue with aluminum wire is that its weak. If you have to 'push' the wire through a conduit and then a troch, as in a MIG welder, then any drag, even the slightest bit, will cause the wire to
bunch up....and you have to reset the wire.

Binzel and TBI have a Mig torch that has a constant torque wire tractor motor in the torch, and this 'pulls' the wire through the conduit and only has to 'push' the last few inches through the
neck/tip and into the weld pool. In New Zealand these torches are worth about $2500NZD....not cheap. These are water cooled.

The Miller XRA torch has a small spool of wire in the handpiece and a variable speed wire feed motor also in the handpiece. The handpiece is bulky as a result but its possible to use small diameter
wire, say 0.9mm and 1mm with care. These torches are over $3000NZD and some models are water cooled.

Cobra, the RollRoyce of torches has a speed controlled tractor motor in the torch and will 'push' the wire out through the neck and tip. They are typically water cooled and are just the ticket for high duty cycle
heavy aluminum welding. These torches are over $4000NZD.

The one thing that ties these torches is that they all have a motor to 'launch' the wire. Without such a motor you have to use big diameter wire, 1.2mm or bigger, otherwise you risk repeat wire foul ups.

In order to feed wire, even on an intermittent basis into a TIG weld pool will require a agile wire feed motor right at or very close to the torch. Using a large diameter wire will increase the stop/start
nature of the feed, whereas is you use 0.9mm wire you may be able to feed continuously.

Craig

[peteeng]

Hi Spot - Get a soft wire. The higher strength ones tend to be stiff and you will have issues in the feeder. Welding mild steel is not very critical and in this application where strength is not important get the most economical "soft" wire. Good luck on the feeder. Maybe quicker just to manually weld with a feeder vs trying to automate the welding. Peter

[G-Spot]

Hi,
I worked for seven years fixing and servicing welding equipment.

Craig

Thanks for the info, unfortunately I am having to use the welder to first weld up the machines frame with structural steel, not Aluminium.

What would be a good wire to TIG weld structural steel I beams to the same grade plate ?

I see the replacement parts for the 3D printers are all upgrades for high speed feeding of flexible plastic. Check the pics out, the yellow one is $75 and the other one $150, you can use them with 1.7mm wire but I think I could modify them to run on a few mm larger. They are all metal and use standard stepper motors so give you complete control of the feeding. The one is a push type, the other pull.

What would be a good wire thickness and type to use in your opinion?

Those pro ones you listed are out of my budget, they would need to take the torch out of my hand and do the whole job for that price...lol

[joeavaerage]

Hi,
to be honest I would not be using a TIG on steel.

I would hire a three phase MIG and use CO₂ shielded flux cored wire. Its superb, makes a beautiful high penetration, high deposition rate weld. It looks very much like sub-arc.....beautiful....
and really flatters your welding......and FAST. In one day you can weld up a mountain.

You want to use 1.6mm or 1.8mm wire (if you can get it), and get right up there, 300A or so at 25V.

The same company I worked for hires welders and for about $200NZD/day I'd get a 300A Lincoln inverter MIG. Wire and gas on top.

Craig

[joeavaerage]

Hi,
if you really want to use TIG, then just get MIG wire....comes on spools ready to go. Why re-invent the wheel?

By the way neither of those feeders you linked to are going to be ANYTHING like rugged enough to 'push or pull' wire.
If you want a wire feeder then get a Miller S54.....no pissing about!. There is the S54....and then there's junk.....no in betweens.

https://www.ebay.com/itm/30547963339...3ABFBMvvza4tRj

Craig

[G-Spot]

Hi Spot - Get a soft wire. The higher strength ones tend to be stiff and you will have issues in the feeder. Welding mild steel is not very critical and in this application where strength is not important get the most economical "soft" wire. Good luck on the feeder. Maybe quicker just to manually weld with a feeder vs trying to automate the welding. Peter

what is the closest diameter to 1.7mm with a soft spooled wire ?

If I use such a small diameter wire welding 15mm thick steel plate then it will just have to feed faster, correct ? or will so much surface area of cold metal getting dumped quickly into the weld pool cool it too much?